| Hematopoiesis is a precisely controlled process, which starts from hematopoietic stem cells (HSCs), gradually loses potential of differentiation, and terminates with mature functional red blood cells (RBCs) by enucleation and formation. Recently, a number of studies on regulation mechanisms of erythropoiesis involving transcription factors, miRNAs and lncRNAs during erythroid development had been performed; however, analyses focusing on the integration of the three types of RNAs are yet to be done. In the present study, we compared the dynamic expression profiles of coding gene, miRNA, and lncRNA, discovered similarities and discrepancies among them, predicted potential RNA interactions by calculating expression correlations, and built a functional network controlling the maturation of red blood cells during terminal differentiation.Distinct stages of terminal erythrocytes were obtained during in vitro induction of HSCs derived from human umbilical cord blood. By detecting expressions of biomarkers, we sorted out samples representing four stages during erythroid differentiation, together with purified HSCs and RBCs. We took advantage of high throughput sequencing technologies to obtain expressed coding genes, miRNAs and lncRNAs. The number of expressed coding genes is larger than non-coding RNAs (ncRNAs), especially lncRNAs showed expression level below 30 FPKM (around 97.7%). Most lncRNAs (about 41.4%) were detected in only one sample. This demonstrated that lncRNAs showed strong stage-specific feature that gave potential to be better biomarkers for differentiation stages.By K-means clustering, we classified differentially expressed genes into four subtypes, and annotated each of them by signaling pathway enrichment analyses. The results showed that the enucleation and deformation of RBCs were companied by inhibiting of P53 signaling pathway and cell cycle pathway. In addition, most known chemokine gens were highly expressed in HSCs, and the chemokine signaling pathway was vital for maintenance of self-renewal and fate decisions of HSCs.We obtained miRNA targets datasets from miRSystem database. By further integrating with expression data of mRNAs and miRNAs, and filtering by Pearson’s product correlation coefficient and P-values, we identified interactions between miRNAs and genes in four subtypes. The results showed that hsa-miR-107 and hsa-miR-548h-5p showed high expression in RBCs, promisingly inhibited the expression of chromatin structure associated genes to promote maturation and enucleation. The potential function of hsa-miR-532-3p was related with promoting translation to accumulate a great amount of protein to maintain the short life of RBCs.Integrating lncRNA expression data with mRNA expressions, we discovered that lncRNAs played key roles in erythroid differentiation, including up-regulating the expression of genes involving reactions to DNA damage stimulus during differentiation while down-regulated them in RBCs, up-regulating the expression of genes involving hematopoiesis in HSCs and RBCs, as well as down-regulating genes encoded cell cycle protein to promote maturation. In addition, we identified potential cis-regulating lncRNAs and their targets and validated expression correlations in K562 cell line. We also identified 83 novel lncRNAs, which enriched the known lncRNA database.To sum up, we built up a regulation network that controlled maturation in terminal differentiation, including interactions between mRNAs and ncRNAs, as well as competitions between miRNAs and lncRNAs. These data provide a substantial resource for the study of normal erythropoiesis, which will permit further investigations and understanding of erythroid development and acquired erythroid disorders. |
PDF Full Text Request